ABSTRACT Recording the electrical impulses of individual neurons in intact brain circuits in real time has been a longstanding goal in neuroscience. One potentially widely applicable use of voltage recording would be to test postsynaptic responses upon physiological or optogenetic activation of presynaptic partners. Recording a neuron while its inputs are controlled would enable a detailed understanding of how individual neurons process information. This understanding becomes important when circuity is altered in disease, e.g. in the striatum with Parkinson's or Huntington's diseases, as it would help explain the pathogenesis of the mutant phenotype and suggest possible therapies. However, the only way now to reliably measure membrane voltage in vivo is to perform electrophysiology, whose difficulty and low throughput hamper widespread adoption. We aim to develop a new paradigm for determining the input-output relationshps of neurons using genetically encoded voltage indicators (GEVIs) and two-photon imaging. GEVIs can provide information on subthreshold voltage changes, which form the basis of neuronal computation and modulate excitability, and on timing and order of neuronal action potentials. These constitute basic essential information required for understanding information processing in brain circuits. However, in vivo voltage imaging is currently limited. No published GEVIs respond with sufficient speed and amplitude for spike detection in single trials under two-photon excitation. In this project, we will develop methods to record electrical activity from individual neurons in the brain at depth using two- photon microscopy. Our approach combines engineering of GEVIs that can respond to two-photon illumination with the establishment of conditions for using GEVIs in brain slices and living brains. Specifically, we will carry out the following aims: (1) Generate brighter and more responsive variants of ASAP2s, the best performer under two-photon excitation, and of Ace-mNeonGreen, a leading performer under one-photon excitation; (2) validate GEVI variants for their ability to report contributions of specific inputs to subthreshold and action potential responses in a variety of neurons of the fly visual system, in single trials, in vivo; and (3) systematically test GEVI performance under two-photon excitation in mouse striatal spiny projection neurons in ex vivo acute brain slices and in living mice in vivo, using GEVIs to determine the role of cell type- specific inputs to a recently discovered phenomenon of long-lasting dendritic voltage plateaus. This project will integrate the expertise of three groups spanning protein engineering, optical method development, and systems neuroscience to improve two-photon imaging of GEVIs so they can be used to image voltage transients in single trials. If successful, this project will open up in vivo two-photon imaging of GEVIs to many interested researchers, potentially catalyzing a transformation in how we measure neuronal responses in living brains.